Download Vacuum Ultraviolet Spectroscopy and Photochemistry of Zinc

Article
pubs.acs.org/JPCA
Vacuum Ultraviolet Spectroscopy and Photochemistry of Zinc
Dihydride and Related Molecules in Low-Temperature Matrices
Chris Henchy, Una Kilmartin, and John G. McCaffrey*
Department of Chemistry, National University of Ireland Maynooth, Maynooth, Co. Kildare, Ireland
S Supporting Information
*
ABSTRACT: Optical absorption spectra of thin film samples,
formed by the codeposition of zinc vapor with D2 and CH4,
have been recorded with synchrotron radiation. With
sufficiently low metal vapor flux, samples deposited at 4 K
were found to consist exclusively of isolated zinc atoms for
both solids. The atomic absorption bands in the quantum
solids D2 and CH4 were found to exhibit large bandwidths,
behavior related to the high lattice frequencies of these low
mass solids. The reactivity of atomic zinc was promoted with
1
P state photolysis leading to the first recording of electronic
absorption spectra for the molecules ZnD2 and CH3ZnH in the vacuum ultraviolet (VUV) region. 3P state luminescence of
atomic zinc observed in the Zn/CH4 system points to the involvement of the spin triplet state in the relaxation of CH3ZnH
system as it evolves into the C3v ground state. This state is not involved in the relaxation of the higher symmetry molecule ZnD2.
Time dependent density functional theory (TD-DFT) calculations were conducted to predict the electronic transitions of the
inserted molecular species. Comparisons with experimental data indicate the predicted transition energies are approximately 0.5
eV less than the recorded values. Possible reasons for the discrepancy are discussed. The molecular photochemistry of ZnD2 and
CH3ZnH observed in the VUV was modeled successfully with a simple four-valence electron AH2 Walsh-type diagram.
I. INTRODUCTION
Zinc dihydride (ZnH2) has been the subject of a number of
investigations in recent years. Two primary reasons can be
identified for the current interest in this triatomic molecule.
The first is that ZnH2 can be considered the parent molecule
for the Group 12 metal M(ns)2 dihydrides and considerable
interest thereby exists for the determination of its fundamental
molecular constants. The second stems from the rich
photochemistry of the related dialkyl metals, MR2, such as
dimethylzinc. Of particular significance in obtaining precise
molecular constants for ZnH2 was the high-resolution mid-IR
emission study conducted by Bernath and co-workers.1 Analysis
of rotationally resolved gas phase spectra for ZnH2 and ZnD2
allowed extraction of the equilibrium bond lengths as 1.52413
and 1.52394 Å, respectively, of these two 64Zn-containing
isotopologues. The band origins of the asymmetric stretching
vibrations for 64ZnH2 and 64ZnD2 were identified at 1889.433
and 1371.631 cm−1, respectively. In addition, verification of the
expected linear, centro-symmetric geometry was also achieved.
Quite recently, Sebald et al.2 have conducted high level ab initio
calculations to generate an accurate potential energy function
for the ground X 1Σg+ electronic state of ZnH2. The coupledcluster method utilized in the theoretical study yielded
molecular constants in excellent agreement with Bernath’s
experimental values for the three hydrogenic isotopologues up
to approximately 15 000 cm−1 above the equilibrium minimum.
Recently microwave absorption spectroscopy3 has been used to
determine the structure and bonding in methylzinc hydride
© 2013 American Chemical Society
(CH3ZnH). The Zn−H bond length was determined to be
1.521 Å, very similar to that of ZnH2. Consistent with the linear
structure of zinc dihydride, the geometry of CH3ZnH was
found to be C3v.
The vibrational spectroscopy of zinc dihydride has been
recorded in a variety of matrix-isolation experiments. Wang and
Andrews4 generated this species with broadband photolysis of
atomic zinc in solid para-H2 (and H2-doped neon) observing
the asymmetric stretching mode at 1880.5 (1880.6) cm−1 and
the bending mode at 631.9 (632.5) cm−1 for 64ZnH2. The two
other hydrogen isotopologues were also reported in that study,
and results were compared with DFT predictions. These modes
were reported previously by Andrews5 as 1870.8 and 628.3
cm−1 for ZnH2 in argon matrices. Our group6 also observed
them at this position in Ar but generated with secondary
photolysis of dimethylzinc.
Quantum chemical calculations have been used at a variety of
levels of sophistication to predict the bond length and
vibrational frequencies of ground state zinc dihydride. A
comparison with Bernath’s experimental values in Table S1
reveals that DFT calculations conducted with the B3LYP
functional and the 6-311++G(3df,3pd) basis set provide results
in good agreement with observations. This basis set has been
used extensively in the present work, yielding results in
Received: May 13, 2013
Revised: August 23, 2013
Published: August 23, 2013
9168
dx.doi.org/10.1021/jp404695a | J. Phys. Chem. A 2013, 117, 9168−9178
The Journal of Physical Chemistry A
Article
complete agreement with the earlier1 DFT findings. Thus the
ground state equilibrium bond length was found to be 1.5413
Å, and the fundamental vibrational frequencies of ν1, the
symmetric (σg), and ν3, the asymmetric (σu) stretches were
1915.5 and 1926.8 cm−1, respectively, while ν2, the bending
(πu) mode, is predicted at 642.8 cm−1 for 64ZnH2. Table S1 in
the Supporting Information collects these DFT harmonic data
and 64ZnD2 results for comparison with a variety of
experimental IR findings both in the gas phase and matrices.
Detailed gas phase collisional deactivation studies have been
conducted by Breckenridge and co-workers7 on the excited
triplet 4p 3P1 state of atomic zinc8 with pump−probe methods,
for molecular hydrogen and its isotopologues, HD and D2.
More recently Umemoto and co-workers9 have extended this
work to the excited singlet 4p 1P1 state. Analysis of the
rotational distribution in the zinc hydride product molecule has
allowed the inference of a long-lived intermediate in these
reactions. This proposal has been examined by Novaro’s
group10 who have determined the reaction paths of excited
state atomic zinc with molecular hydrogen. The findings of this
theoretical work largely confirm the existence of a deeply
bound linear inserted H−Zn−H species, but involving an
endothermic reaction with respect to the separated species
ground (1S0) state atomic zinc and molecular hydrogen.
In the present contribution we examine the electronic
spectroscopy of the zinc dihydrides (ZnD2 and ZnH2) as well
as methylzinc hydride (CH3ZnH) generated from the reaction
of excited 1P1 state atomic zinc with D2/H2 and CH4. This is
achieved for atomic zinc isolated in solid matrices formed from
the co-condensation of zinc vapor with molecular deuterium
(D2), methane (CH4), and H2 in doped Ar matrices. Any
luminescence detected in the current reactive host solids as a
result of zinc 1P1 photoexcitation is compared with that
previously observed by our group for atomic zinc isolated in
inert solids such as neon11 and argon12 matrices. The electronic
spectra of the resulting “inserted” species have been recorded
for the first time in the vacuum-UV range. Furthermore,
repeated scans in the VUV region have revealed photodissociation of the inserted molecular species regenerating
atomic zinc. Time-dependent density function theory (TDDFT) calculations were then used to analyze the nature of the
electronic transitions responsible for the absorption spectra of
the inserted molecules. This method was also used to examine
the efficient photodissociation observed for these species in the
VUV.
a 0.25 m Spex 240 M imaging monochromator which was
coupled by a 2 m optical fiber to the sample. The CCD chip
consists of 1100 × 330 pixels and was operated at −100 °C
with exposure times typically of 5 min.
Samples were made by the codeposition of zinc vapor with a
large excess of the host gas (deuterium or methane) onto an
LiF window at 4 K. This was the minimum temperature
attainable on the sample holder, with the Leybold−Heraeus
flow-through liquid helium cryostat used. All of the host
materials used in the present study, Ne, Ar, D2, H2, and CH4,
are transparent in the wavelength range of interest with the
exception of CH4, which starts to absorb for λ < 145 nm. Zinc
vapor was produced by electron bombardment of a 2.0 mm Zn
rod placed in a molybdenum holder. Full details of the electron
bombardment metal atom source used have been presented
elsewhere.13 In contrast to solid D2, none of the attempts made
at 4 K to isolate zinc atoms in neat, solid hydrogen were
successful. However, Ar samples containing up to 10% H2 were
found to be stable even when deposited at 12 K. Zn/H2
photochemistry was analyzed in these H2-doped Ar matrices.
B. Calculations. The DFT calculations conducted in this
study all used the B3LYP functional and were performed with
the Gaussian-03 suite of programs14 running on a Linux
workstation, which has been described15 in our earlier work. In
the present study the 6-311++G (3df,3pd) basis set was used
exclusively. As indicated in Table 1, this method yielded an
optimized geometry and vibrational frequencies of ZnH2
identical to the values previously published in Table S13 of
the supplementary data of ref 1. Moreover, the DFT
predictions obtained for CH3ZnH also agreed with earlier
DFT,16 microwave,3 and matrix results5 as listed in Table S1 of
the Supporting Information (SI).
Table 1. A Comparison of the Structural and Vibrational
Data of CH3−Zn-H Predicted by the Current DFT
Calculations and Determined in Previous Experimental
Studiesa
DFT/B3LYP
Zn−H
Zn−C
C−H
Zn−C−H
H−C−H
H−Zn−C
II. METHODS
A. Experimental Section. Absorption spectra in the
vacuum ultraviolet (VUV) region were recorded using
synchrotron radiation with the HIGITI setup on the W3.1
beamline at HASYLAB, DESY, Hamburg, Germany. Scans in
the 120−350 nm range were done with an LiF filter in the
primary monochromatora modified 1.0 m Wadsworthto
preclude short wavelength light reaching the sample via high
order transmittance. A sodium-salicylate converter was used
with an XP2020 photomultiplier tube (PMT) for light
detection in the VUV. Emission spectra in this region were
recorded with a 0.4 m Seya-Namioka vacuum monochromator
using a Hamamatsu (model 1645U-09) multichannel plate for
photon detection. A liquid nitrogen-cooled, charged coupled
device (CCD) detector (Princeton Instruments, model LN/
CCD 1108PB) was used to record emission spectra in the UV−
visible region. This steady-state CCD camera was mounted on
C−H s str
H−Zn str
CH3 s bend
C−Zn str
C−H as.str
CH3 as bend
CH3 rock
C−Zn−H bend
experimental results
Bond Length (Å)
1.5425
1.9546
1.0912
Bond Angle (deg)
110.9
107.9
180.0
Frequencies (cm−1), A1 modes
3021.1
1902.7
1208.4
552.3
Frequencies (cm−1), E modes
3094.2
1456.4
702.9
422.9
1.521
1.928
1.14
110.2
180.0
2920
1866
1179
564
687
443
a
The structural experimental data are from gas phase microwave
spectra presented in ref 3, while the vibrational data are from matrix
results presented in ref 5. The DFT vibrational frequencies quoted are
harmonic values. All of the vibrational frequencies listed are in
wavenumber (cm−1) units, while the bond lengths are in Angstrom
units.
9169
dx.doi.org/10.1021/jp404695a | J. Phys. Chem. A 2013, 117, 9168−9178
The Journal of Physical Chemistry A
Article
dashed vertical line in Figure 1, a blue shift of the band occurs
from the 213.9 nm gas phase position17 of the atomic
transition. Two other major characteristics of the atomic
absorption profile recorded in solid D2 are the large bandwidth
(∼2050 cm−1) and the pronounced 3-fold splitting with
resolved peaks at 206.4, 209.6, and 213.1 nm.
Shown by the middle trace in Figure 1 is the absorption
spectrum recorded for zinc isolated in a pure methane matrix
(Zn/CH4). In contrast to the deuterium sample, the absorption
band is now centered slightly to the red of the gas phase
transition and is considerably narrower. However, it also
exhibits characteristic 3-fold splitting, with partially resolved
features at 212.6, 214.5, and 216.5 nm. For comparison
purposes, the absorption spectrum of atomic zinc isolated in
solid neon11 is shown by the bottom trace in Figure 1. While
the atomic band in neon is located blue of the gas phase
transitionroughly in the same region as Zn/D23-fold
splitting is not evident and additional bands due to Zn dimer
are now present at the indicated positions. The existence of
pronounced dimer bands in neon samples reveals more efficient
isolation of zinc atoms in solid deuterium and methane
matrices than in neon. This difference will be discussed in more
detail ahead in relation to possible site occupancies.
Lineshape analyses done on the absorption bands of the Zn/
D2, Zn/CH4, and Zn/Ar systems are presented together in
Figure S1 of the Supporting Information. In the Gaussian fits
shown in Figure S1 three components, all with the same line
width, were used on a given system. With the exception of Zn/
CH4, where a minor blue site exits, the use of only three
components provides a good description of the recorded bands.
The same fitting procedure was successful for all of the systems,
the details of which are collected in Table 2. This indicates that
similar 3-fold splitting of the 4p 1P1 ← 4s 1S0 electronic
transition occurs in the three hosts. As the Zn/Ne absorption
and excitation spectra have recently been analyzed11 in detail,
only the relevant aspects will be compared with the present
data. The fits done on Zn/Ne utilized different parameters,11
including variable linewidths and the requirement of a fourth
Excited state calculations on ZnH2 (CH3ZnH) were carried
out with the TD-DFT method using the optimized geometry of
the D∞h (C3v) ground state molecule. Once calculations had
converged, the excitation energies, oscillator strengths (f), and
transition dipole moments of the first 50 excited states were
extracted. The energies of the molecular orbitals were obtained
using the “pop=reg” command, yielding values for all 16
occupied and 101 virtual orbitals of ZnH2. Excited state
calculations on methylzinc hydride (CH3ZnH) were conducted
in a similar fashion but in this case involving 20 occupied and
172 virtual orbitals. The molecular orbital energies were also
determined as a function of angle between the linear and bent
(90°) geometries. By doing these calculations in 5° increments,
so-called “Walsh” diagrams were thereby produced. The MO
plots shown in these diagrams were produced from the
electronic density. This was done in Gaussian 03 by generating
a “cube” file for a given molecular orbital. The electronic
density was then plotted as a wireframe over the molecule
showing the shape and parity of the orbitals of interest.
III. RESULTS
A. Deposition. An absorption spectrum recorded in the
VUV-UV region for a Zn/D2 sample formed with low metal
loading is shown by the top trace in Figure 1. Only a single,
albeit structured, band is present whose center feature is located
at 209.6 nm. Due to the sparsity of electronic transitions of
atomic zinc in the wavelength range longer than 200 nm, this 3fold split absorption band can immediately be assigned to the
resonance 4p 1P1 ← 4s 1S0 transition. As indicated by the
Table 2. Absorption Band Positions Recorded for the 4p
1
P1−4s 1S0 Transition of Atomic Zinc Isolated in the Solids
Formed from the Light Materials Deuterium, Methane,
Neon, and Argon at 4.2 Ka
Zn
D2
CH4
Ne
Ar
Figure 1. Absorption spectra recorded for Zn isolated in the molecular
matrices D2 and CH4 deposited at 4 and 12 K, respectively, both of
which were scanned at 4.2 K. The location of the resonance 4p 1P1 ←
4s 1S0 transition of atomic zinc in the gas phase (213.9 nm) is
indicated by the dashed vertical line. As indicated in the figure, a blue
shift of the entire band occurs for Zn/D2 from the gas phase position,
but a red shift exists in Zn/CH4. For comparison the absorption
spectrum of Zn/Ne is presented on the bottom revealing a less
resolved main band and features recently attributed (ref 11) to zinc
dimer.
absorption center (nm)
absorption energy (cm−1)
Δ (cm−1)
206.4
209.6
213.1
212.6
214.5
216.5
203.68
205.17
206.15
205.15
206.75
208.47
48444
47700
46992
47025
46614
46191
49095
48740
48509
48747
48367
47967
742
742
742
435
435
435
600
561
375
412
412
412
The gas phase position17 of this transition is at 46745.413 cm−1. All
of the matrix energies quoted are in wavenumber (cm−1) units. The
bandwidths (Δ, fwhm) quoted for the matrix features were extracted
in the Gaussian lineshape fits of the spectra presented in Figure S1.
The quoted values were calculated as Δ = s√[8 ln(2)] where s is the
variance in the fitted function, G(x) = a exp[−(x − m)2/(2s2)].
a
9170
dx.doi.org/10.1021/jp404695a | J. Phys. Chem. A 2013, 117, 9168−9178
The Journal of Physical Chemistry A
Article
component. This contrasting behavior points to different site
occupancy in solid neon compared with the molecular hosts
studied in the present work and the earlier12 Ar data.
B. Luminescence. No emission was observed for Zn/D2
samples either in the UV or in the visible spectral regions with
excitation into the atomic resonance absorption band shown in
Figure 1. This situation is depicted in the central panel of
Figure 2 with excitation at 210 nm. It is in stark contrast to
Figure 2. A summary of the emission spectra recorded at 4.2 K in the
Zn/D2 and Zn/CH4 systems. As indicated by the middle trace, no
emission exists in the Zn/D2 system, while only the spin triplet is
present (bottom trace) for Zn/CH4. It is located at the red-shift of the
atomic zinc 4p1 3P1−4s 1S0 transition which occurs in the gas phase at
approximately 308 nm. For comparison the emission recorded in Zn/
Ne is shown in which singlet state atomic emission is observed.
Figure 3. Changes observed in the VUV−UV absorption spectra of
the Zn/D2 system upon atomic photolysis at specific wavelengths
(indicated by the arrows in Figure 1) shown as difference spectra. The
negative peaks in the plots indicate the loss of the atomic bands, while
the positive ones reveal the growth of transitions in the VUV. In
addition to the pure D2 matrix result, the response of a 10% D2 in Ar
sample and a 10% H2 in Ar sample to similar irradiation are shown in
the lower panel. The dip evident at approximately 185 nm in the 10%
D2 in Ar sample arose due to a change in the SR beam position during
the scan and is not of photochemical significance.
atomic zinc isolated in the solid methane where, as shown in
the bottom panel, singlet excitation at 214 nm produces
emission at 348 nm and a red shoulder at 380 nm. From the
long-lived nature18 of this near-UV emission, these bands can
be attributed to the atomic Zn 4p 3P1 state which occurs in the
gas phase17 at 307.68 nm.
Comparable excitation at 205.4 nm in Zn/Ne produces11
intense emission at 212.8 nm as shown in the top trace of
Figure 2. With a radiative lifetime of 1.16 ns and its spectral
location, these characteristics indicate that this emission can be
assigned to the fully allowed 4p 1P1 → 4s 1S0 transition. Thus,
singlet P-state fluorescence dominates in neon,11 as it does in
argon and krypton matrices.12 In light of the data presented in
Figure 2, it is thereby evident that singlet fluorescence of atomic
zinc is completely quenched in both solid D2 and CH4. In the
latter system, the atomic triplet state is however produced with
high efficiency, a situation observed previously in Zn/Xe.
C. Atomic Photolysis. The reactivity of excited 4p 1P1 state
atomic zinc in solid deuterium was monitored in the VUV upon
photolysis at 210 nm, a wavelength indicated by the upper
arrow in Figure 1. The result of a 15 min irradiation is
presented as a “difference” absorption spectrum in the top trace
of Figure 3. In this spectrum depletion of the 3-fold-split atomic
resonance band centered at 210 nm is very evident, but more
significantly, this depletion is accompanied by the appearance
of a pair of broad, featureless bands at 145 and 155 nm as well
as the hint of a much weaker band at 168 nm.
Because of the absence of samples composed of atomic zinc
isolated in neat H2 matrices, the photochemical activity of zinc
with hydrogen was examined in Ar samples containing 10% H2
(Zn/H2Ar). To allow direct comparison with the results
obtained for the pure Zn/D2 system, D2-doped Ar samples of
composition similar to H2Ar were also produced. The effect of
atomic zinc photolysis in a Zn/D2Ar sample is shown in the
middle trace of Figure 3. As observed with pure D2 samples
(upper trace), new absorption bands are produced in the 140−
160 nm region, but the changes produced in the atomic
absorption in the doped D2/Ar samples are more complex. In
particular, an increase is observed in one feature located at 215
nm which is due to photolytically induced site interconversion,
that is, the isolation of the zinc atom in a new site in the D2doped Ar sample. A similar, but more pronounced, effect is
evident in the difference spectrum recorded for a Zn/10%H2Ar
sample, shown in the bottom trace in Figure 3. This effect arises
from ill-defined sites of isolation of atomic zinc in the lessstructured doped-Ar samples. Clearly the VUV absorptions of
Zn/10%H2Ar occur in the same spectral region but are less
structured than in the corresponding D2-doped Ar samples. It is
also evident that the amount of photochemical product
generated in the H2-doped Ar samples is less than in the
equivalent D2-doped Ar samples. This is probably related to the
less efficient trapping of the lighter isotopologue in Ar.
Excited state reactivity of zinc with methane was examined
with photolysis at 214 nm whose location is shown by the
9171
dx.doi.org/10.1021/jp404695a | J. Phys. Chem. A 2013, 117, 9168−9178
The Journal of Physical Chemistry A
Article
photolysis. As indicated in the two lower traces in Figure 5,
irradiation at 145 and 153 nm into each excited molecular band
arrow in the middle panel of Figure 1. The result of a 15 min
irradiation at this wavelength is presented as a “difference”
absorption spectrum in the upper trace of Figure 4. In this
Figure 5. VUV photodissociation of ZnD2 isolated in solid D2
revealing, in the lower two traces, the regeneration of atomic zinc
and the depletion of the molecular bands.
Figure 4. Same as for Figure 3 except for the Zn/CH4 system. The
lower trace shows the results obtained for a 5% methane in the Ar
sample.
resulted in the depletion of both. This depletion was
accompanied by the regeneration of the characteristic atomic
zinc absorption at 210 nm. After regeneration of atomic zinc,
the production of the inserted molecule was shown to be
possible once again with atomic photolysis. Thus the insertion/
dissociation cycle could be repeated several times. As a result of
this and the characteristic atomic zinc absorption band shape, it
is highly likely the other product of the photodissociation is
molecular hydrogen and not hydrogen atoms. At the very low
temperatures this work was done, some isolation of H atoms
should be possible in the solid hosts. The absence of H atoms is
indicative of a concerted dissociative mechanism whereby both
Zn−H bonds break at the same time, leading to the formation
of an H−H bond.
In the Zn/CH4 system all three VUV molecular bands were
found to be photochemically active. Irradiation into any of
them resulted (data not shown) in the regeneration of the
atomic zinc absorption at 214 nm and depletion of the new
molecular bands. This behavior is similar to that observed by us
previously19 in the photochemistry of matrix-isolated dimethylzinc (DMZ). However, in contrast to the dissociation of
DMZ, where secondary photolysis never regenerated the
starting material, methylzinc hydride species could be
regenerated with atomic photolysis. This difference stems
from the inability of metal atoms to insert into the C−C bond
of ethane compared to their facile insertion into the C−H
bonds.
E. TD-DFT Calculations. The electronic absorptions of
ZnH2 predicted by TD-DFT calculations are collected in Table
3 for the singlet−singlet transitions. In this table it is evident
that the lowest energy transition is from the HOMO (MO 16)
spectrum depletion of the atomic resonance band is evident as
well as the appearance of at least three featureless bands at 174,
162, and 149 nm. These bands are all in the vacuum-UV but at
lower energy than those observed in Zn/D2. The noise evident
at wavelengths shorter than 145 nm in the Zn/CH4 spectrum is
due to the strong absorption of the host material, methane, in
this region.
The reactivity of excited state atomic zinc with methane was
also examined in CH4 doped-Ar matrices. The “difference”
absorption spectrum shown by the bottom trace in Figure 4
presents the results of a 15 min irradiation at 208 nm. In this
5% CH4/Ar sample, depletion of the atomic resonance band is
much simpler than what is observed in the H2-doped Ar
samples with clean loss of the atomic features, that is, free from
site interconversion. The appearance of at least three featureless
bands in the VUV is also evident in the figure with a better S/N
ratio than obtained in the pure methane sample. However, as
shown by the gray trace in the upper panel of Figure S2, the
three new bands in Ar are all to the blue of the three bands in
pure methane (black trace). The locations of the bands in
doped-CH4 Ar samples are 168, 156, and 146.5 nm compared
with 174, 162, and 149 nm in neat CH4. In contrast, the bottom
panel of Figure S2 reveals that little or no shift occurs in the
bands observed in the Zn/D2 system in pure deuterium (red
trace) and D2-doped Ar matrices (gray trace). The contrasting
behavior of the VUV bands of ZnD2 and CH3ZnH in Ar may
be related to site occupancy whereby the latter is cramped in a
divacancy, while the former fits easily into such a site.
D. VUV Photolysis. Both of the strong vacuum-UV bands
in the Zn/D2 system were found to be very sensitive to
9172
dx.doi.org/10.1021/jp404695a | J. Phys. Chem. A 2013, 117, 9168−9178
The Journal of Physical Chemistry A
Article
Table 3. Excitation Energies and Oscillator Strengths for Spin-Allowed Transitions from the Ground X 1Σg State of ZnH2a
state no.
1
2
3
4
5
6
7
8
9
10
11
orbitals/coeff.
16
16
16
16
16
16
15
15
15
15
15
16
16
16
15
16
15
15
15
15
15
15
16
16
→
→
→
→
→
→
→
→
→
→
→
→
→
→
→
→
→
→
→
→
→
→
→
→
17/0.31058
18/0.62335
17/0.62335
18/−0.31058
19/0.66612
20/−0.23178
17/0.10422
18/0.66126
17/0.66126
18/−0.10422
21/−0.10725
19/0.21522
20/0.58930
26/−0.18648
20/−0.10301
21/0.68731
19/0.25477
20/0.66992
21/0.10189
21/−0.40793
22/0.56005
21/0.56142
20/0.13779
22/0.35017
state sym.
E (eV)
λ (nm)
E (cm−1)
f
Πg
5.7997
213.78
46777
0.0000
Πg
5.7997
213.78
46777
0.0000
Σu
7.2816
170.27
58730
0.0910
Πu
8.0394
154.22
64842
0.3502
Πu
8.0394
154.22
64842
0.3502
Σu
8.1768
151.63
65950
0.7040
Σg
8.3336
148.78
67213
0.0000
Σg
Σg
9.0256
9.3425
137.37
132.71
72796
75352
0.0000
0.0000
Σu
9.8924
125.33
79789
0.0217
Σu
10.0716
123.10
81234
0.0003
1
1
1
1
1
1
1
1
1
1
1
Only transitions up to approximately 10 eV are provided. Bold font indicates the first allowed transition at 170.3 nm from the HOMO to the
LUMO+1 orbital and the mixture of three electronic transitions that make up the strongest band (see text for details).
a
to a degenerate LUMO pair of orbitals (MOs 17 and 18)
corresponding to the A1Πg ← X1Σg+ state transition. This is
what is predicted by a qualitative AH2 Walsh-type20 diagram for
the simplest four-valence electron system. However, from the fvalues provided in Table 3, it is clear that while the lowest
energy transitions occur at long wavelengths (in the region of
the atomic zinc 4p 1P1 ← 4s1S0 transition) they have no
oscillator strength due to being g−g forbidden. Consistent with
experimental observations, the first transitions are predicted in
the TD-DFT calculation to occur in the VUV. The calculated
absorption of ZnH2 is shown by the stick spectrum in the lower
panel of Figure S3. Shown also is the spectrum convoluted with
a Gaussian function having a line width of 800 cm−1. As
indicated in Table 3, the first allowed transition at 170.3 nm is
from the HOMO to the LUMO+1 orbital. The strongest band,
with a convoluted center at approximately 154 nm is, as
specified in Table 3, a mixture of three electronic transitions. It
consists of a degenerate 1Π state at 154 nm and a 1Σu state at
151.6 nm. The apparent disagreement between the intensities
of the stick and convoluted spectra at 154 nm is an artifact of
the presentation which arises due to the degeneracy of the
excited 1Π states at this location.
Shown in the bottom panel of Figure 6 are the absorption
spectra recorded for ZnD2 both in a neat D2 solid (red curve)
and in a D2-doped Ar matrix (gray curve). By comparing these
curves with the convoluted DFT prediction21 (black curve, in
the upper panel), it is evident that while the predicted bands are
approximately in the same spectral region (the discrepancy is
4200 cm−1) the shapes are quite different. In particular a strong
band consisting of two unresolved components of roughly
equal intensity and a much weaker red band are predicted,
while a broad structure consisting of a dominant band and a
strong red shoulder are observed (red/gray traces in the lower
Figure 6. Comparison of the TD-DFT predicted (convoluted)
excitation energies with the absorption spectra of ZnD2. The dashed
curves in the lower panel are those obtained by fitting four Gaussian
curves to the recorded data. Their sum is shown by the gray curve
overlaid on the experimental data, shown in red. The solid black traces
in the upper panel are the stick and convoluted predicted spectra. To
obtain a match with the recorded data, the gray traces were generated
by blue shifting the original TD-DFT bands by 4200 cm−1 (shown by
the solid black trace) and splitting the degenerate Π state by 2200
cm−1. The curve generated by summing the dashed curves is shown by
the solid gray trace, and it can be compared with the experimental data
in the lower panel.
9173
dx.doi.org/10.1021/jp404695a | J. Phys. Chem. A 2013, 117, 9168−9178
The Journal of Physical Chemistry A
Article
doped-Ar system. However, as was observed in the ZnH2
system the predicted bands are all at lower energy than the
recorded bands. In the comparison shown, the intensities of the
TD-DFT calculated bands have not been altered. Other details
of the fit are provided in the caption of Figure 7.
panel). From the TD-DFT calculation it is known (see Table
3) that the transition at 154 nm is to a 1Π state, and it has been
observed that the degeneracy of such states is often removed by
the symmetry of the site accommodating the guest molecule in
the solid matrix.
In an attempt to account for the observed bands, the TDDFT transitions are all blue-shifted by 4200 cm−1, and the 1Π
state has been split by 2200 cm−1, the result of which is shown
by the four dashed curves in the upper panel of Figure 6. In so
doing, a line width of 1050 cm−1 was used for each of the
bands, and for all of them, a constant shift was applied.
However, the predicted transitions intensities were not adjusted
from the f-values obtained directly from the TD-DFT
calculation and listed in Table 3. Addition of these four
components (dashed curves) generates the solid black trace
shown overlaid on the recorded ZnD2 spectra in the lower
panel of Figure 6. The agreement is quite good; however, one
aspect made evident in the comparison shown in Figure 6 is the
existence of a lower energy band of ZnD2 at approximately
60 000 cm−1. This feature is not immediately obvious in the
experimental spectra, in part due to the weakness of this band
but also the poor signal-to-noise ratio.
In Figure 7 the absorption spectra of CH3ZnH recorded in
neat methane (red trace) and in CH4-doped Ar (gray trace)
IV. DISCUSSION
A. Metal Atom Spectroscopy. The absorption bandwidth
(total fwhm 2100 cm−1) of the atomic 4p 1P1 ← 4s 1S0
transition in solid D2 is, as depicted in Figure 1, considerably
greater than that of atomic zinc in the other light solids
examined, namely, methane, and neon. Indeed it is much
greater than what has been observed12 for the rare gases, Ar, Kr,
and Xe. However, it is in line with values of 1500 and 1200
cm−1 measured for atomic magnesium22 and lithium,23
respectively, in solid hydrogen. On the other hand, the 3-fold
splitting recorded in the Zn/D2 system is much better resolved
than in the previous work done on Mg in H2, HD, or D2. From
the spectra recorded for Li23 isolated in the solid hydrogens, it
is not clear that 3-fold splitting exists in this system. The reason
for the difference between zinc and the other metal atom
systems is likely related to the smaller van der Waals radius of
atomic zinc in the Zn·RG diatomics.
Since the zinc atom−hydrogen molecule (Zn·D2) van der
Waals distance is not yet known, one cannot definitively
confirm this proposal. However, spectroscopic work on 1:1
metal atom/rare gas atom van der Waals complexes has
revealed that, of the metals in question (Li, Mg, and Zn), the
zinc-containing species have the shortest bond lengths. Given
that the nearest neighbor (nn) distance24 in solid H2 is 3.79 Å,
considerably greater than the nn distance in solid neon (3.09
Å) but very similar to that of solid Ar (3.76 Å), it is reasonable
to propose that the same site occupancy will occur in solid H2
as solid argon. The results of a combined experimental/pairpotentials25 simulation26 suggest occupancy of atomic zinc in
single vacancy (substitutional) site of argon. Thus, it is plausible
that atomic zinc resides in a substitutional site in D2. This
proposal is supported by the similarity of the 3-fold splitting
revealed in the line shape fits presented in Figure S1 for Zn/Ar
and Zn/D2 matrices. The much greater bandwidth of the latter
system (742 cm−1) compared with Ar (412 cm−1, see Table 2)
arises from the high phonon frequency of D2 which is a lowmass, quantum solid.
As a result of the foregoing considerations of the size of
atomic zinc, it is likely the site accommodating this guest is
more symmetric than the distorted lattices accompanying
isolation of the larger atoms Li and Mg if substitutional
occupancies are also present for these guests. Moreover, the
proposal of this site occupancy is consistent with the much
more efficient isolation of atomic zinc in solid D2 than in solid
neon as revealed in the absorption spectra shown in Figure 1.
Multivacancy occupancy has been proposed11 for the site of
isolation of atomic zinc in solid neon.
B. Zinc Dideuteride Spectroscopy. The most likely
candidate for the VUV absorption bands (shown in Figure 3)
produced after atomic zinc photolysis in solid deuterium is zinc
dideuteride, ZnD2. This molecule has been definitively
identified in FTIR work5 of matrix-isolated atomic zinc in
rare gases containing small amounts of hydrogen (and its
isotopologues). The results obtained by Wang and Andrews4
for the reaction of atomic zinc in the solid para-hydrogen (and
solid ortho-deuterium) confirm these assignments. The
proposed mechanism for the formation of the ZnD2 molecule
Figure 7. Similar comparison to Figure 6 but for CH3ZnH results.
Three bands are sufficient to provide an acceptable fit of the recorded
absorption bands. No splitting was used on the degenerate E state as it
already has cylindrical symmetry in this C3v molecule. A constant shift
does not allow the three convoluted bands to match the three
recorded bands. In this system, a blue shift of 5200 cm−1 lines up the
central band in the pure CH4 system. However, it is evident that this
constant shift underestimates the location of the red band and
overestimates the blue band.
matrices are compared with a TD-DFT calculated electronic
spectrum. When a comparison of both is made, it is evident that
there is reasonable agreement between the broadened TD-DFT
spectrum (black trace in upper panel) and the experimental
spectrum. Thus three well-resolved peaks are predicted
consisting of a strong blue band and two weaker red bands
of equal intensity. As is evident in the lower panel of Figure 7,
the recorded bands show this behavior as best illustrated in the
9174
dx.doi.org/10.1021/jp404695a | J. Phys. Chem. A 2013, 117, 9168−9178
The Journal of Physical Chemistry A
Article
Table 4. Excitation Energies and Oscillator Strengths for Spin Allowed Transitions from the Ground X 1A1 State of CH3ZnHa
state no.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
a
orbitals/coeff.
20
20
20
20
19
20
19
20
19
20
19
20
19
20
19
20
20
17
18
19
19
19
20
20
17
18
20
17
18
20
17
18
20
20
→
→
→
→
→
→
→
→
→
→
→
→
→
→
→
→
→
→
→
→
→
→
→
→
→
→
→
→
→
→
→
→
→
→
21/0.69950
22/0.69950
23/0.68162
24/0.70006
22/−0.13028
26/0.69002
21/−0.13030
25/0.69001
22/0.65977
26/0.13528
21/0.65977
25/0.13530
23/0.48778
27/0.47850
23/0.47674
27/−0.43952
28/0.20483
22/0.30565
21/0.30487
23/−0.10968
24/0.31292
27/−0.12589
28/0.37893
29/0.17783
21/0.10735
22/0.108.19
31/0.68659
22/0.10735
21/−0.10840
30/0.68661
21/0.49798
22/−0.49783
28/−0.32111
29/0.61493
E (eV)
λ(nm)
E (cm−1)
f
E
1
E
1
A1
1
A1
1
A
5.6516
5.6516
6.2587
7.1196
7.6537
219.38
219.38
198.10
174.15
161.99
45583
45583
50479
57421
61732
0.0006
0.0006
0.1425
0.1784
0.0001
1
E
7.6537
161.99
61732
0.0001
1
E
7.9604
155.75
64205
0.2562
1
E
7.9604
155.75
64205
0.2562
1
A1
7.9907
155.16
64449
0.0603
1
A1
8.0550
153.92
64968
0.0846
1
A1
8.8590
139.95
71454
0.0002
1
E
8.8887
139.48
71695
0.0117
1
E
8.8887
139.478
71695
0.0117
1
A2
8.9005
139.30
71787
0.0000
1
A1
8.9127
139.11
71885
0.0467
state sym.
1
Only transitions up to approximately 9 eV are provided.
is the insertion of the excited 1P1 state Zn atom into the D−D
bond and the stabilization of the resulting intermediate in the
low temperature solid. Gas-phase pump−probe studies9,27 of
this reaction propose a similar species in the process of forming
Zn−H + H. Moreover, the linear triatomics ZnH2 and ZnD2
have indeed been generated in the gas phase in a hightemperature furnace-discharge as observed1 in emission in the
mid-IR region.
The absorption bands of zinc dideuteride, ZnD2 (Figure 3)
and methylzinc hydride, CH3ZnH (Figure 4), are featureless in
the VUV spectral region. This characteristic arises because the
excited states involved have all been shown to be dissociative.
As indicated in Figures 6 and 7, the TD-DFT predicted spectra
exhibit the right number of bands which are approximately in
the correct spectral region for both species. However, they do
require blue shifting of about 0.5 eV to achieve numerical
agreement with the recorded data. The reason for this shift
appears to be different for the two species. Thus, in CH3ZnH, a
part of this effect arises from a matrix shift in the experimental
spectra. This is revealed in the comparison of the pure methane
data and the doped methane-Ar results, presented in the upper
panel of Figure S2, which indicates the latter is blue-shifted
against neat methane. The matrix shift arises due to the
different interaction strengths between the guest species and
the two host materials. As the site sizes and host polarizabilities
of solid CH4 and Kr are very similar, the blue shift in Ar would
indicate a repulsive interaction of the excited state guest
molecule in its solid environment. This in turn would suggest
CH3ZnH resides in a cramped site in argon compared with
methane.
In ZnD2 by contrast, the matrix shift appears to be
nonexistent (see the lower panel of Figure S2). Even so, the
theoretically predicted bands are red-shifted by about 4000
cm−1 (Figure 6). A possible reason for the difference is that the
large Franck−Condon (FC) factors which accompany these
dissociative electronic transitions are not included in the
absorption band profiles generated in the present TD-DFT
calculations. With the ultracold conditions used to record the
absorption spectra, large FC factors will extend the TD-DFT
calculated spectra to the blue-shift of the calculated vertical
transition energies. The simple Gaussian convolutions done in
the present simulations are evenly distributed around the
electronic band origin and do not address this FC effect.
Another possible reason for the 0.5 eV discrepancy could be
due to inherent errors28,29 in the TD-DFT method itself. A
rough indicator of the inaccuracy of a TD-DFT calculation is
provided by the negative of the highest occupied Kohn−Sham
(KS) orbital energy. Thus when the calculated excitation
9175
dx.doi.org/10.1021/jp404695a | J. Phys. Chem. A 2013, 117, 9168−9178
The Journal of Physical Chemistry A
Article
Figure 8. A Walsh diagram calculated with the DFT method for HZnH. In performing this calculation, the Zn−H bond lengths were optimized for
each specific angle. The four MO’s shown present the shapes of the HOMO and LUMO for the linear and bent (90°) geometries.
Figure 9. A Walsh diagram calculated for CH3ZnH with the DFT method as a function of the C−Zn−H angle. In doing this calculation, all of the
bond lengths were optimized for each specific angle.
9176
dx.doi.org/10.1021/jp404695a | J. Phys. Chem. A 2013, 117, 9168−9178
The Journal of Physical Chemistry A
Article
shape of the energy curve of the stabilized LUMO (MO17) of
ZnH2 (CH3ZnH, MO21) becomes almost linear after only a
small (approximately 15°) angle deviation from 180° while the
shape of the destabilized HOMO energy, MO16 (MO20), is
curved over the entire angle range. As Figures 8 and 9
respectively reveal, ZnH2 and CH3ZnH both show this same
behavior. The curved behavior of the HOMO energy can be
traced back to the angle dependence of the overlap between the
Zn atom p-orbital and the H-atom s-orbital. As such, it will
simply exhibit a cosine dependence as is well-illustrated in
Figure S4. By contrast the cosine dependence of the LUMO is
limited to small angle deviation from the linear geometry.
Thereafter, the energy of this orbital is stabilized by the
formation of the new H−H bond. This contributing effect is
usually not considered in textbook descriptions of AH2 Walsh
diagrams.
energies approach this threshold, the predicted values are likely
to be underestimated relative to experimental results. Following
this rough accuracy guide, the negative of the HOMO energy of
ZnH2 (CH3ZnH) is 63445.7 (57855.7) cm−1, and the first
allowed transition is, as listed in Tables 3 and 4, less than this,
having a value of 58730 (50479) cm −1 . From these
comparisons we expect that, while not an ideal situation, the
B3LYP functional (with the large basis set used) should allow
the TD-DFT method to provide a reasonable prediction of the
excitation energies. Indeed, the underestimate of the predicted
transition energies is less than 6% (0.5 eV) on the allowed
transitions in the 8 eV region. It should be mentioned that the
TD-DFT method also underestimates the atomic zinc 4p 1P1
← 4s 1S0 transition but by a somewhat smaller amount. Thus
the gas phase value is known17 to be 46745.413 cm−1, while the
predicted value is 45531 cm−1corresponding to an underestimate of 2.6%.
C. Photochemistry. The dissociative activity exhibited by
the VUV bands of ZnD2 and CH3ZnH upon photolysis is
consistent with the featureless nature of these absorption bands.
This behavior can be related to the bonding described for a
four-electron AH2 system by a Walsh diagram. Thus the ground
state is, as observed experimentally, predicted to be linear in
this simple bonding model. The first excited states on the other
hand are expected to be bent due to strong stabilization of the
LUMO orbital upon bending.
To investigate the observed photodissociation of ZnH2 and
CH3ZnH, we explored the evolution of the orbital energies of
both molecules from the linear to bent geometries. In analyzing
this we present in Figures 8 and 9 graphs of the ZnH2 and
CH3ZnH MO energies between 180° and 90°, calculated in 5°
increments. Also included are the MO shapes as determined in
Gaussian 03. From Figure 8, it can be seen that when ZnH2 is
in the ground state and the HOMO is occupied, the linear form
is more stable than the bent. When it absorbs at 154 nm, one of
the valence electrons in MO15 (HOMO-1) is promoted to the
LUMO (MO’s 17/18) as indicated in Table 3. In the excited
state it can be seen that the bent form of ZnH2 is actually more
stable than the linear. The angle dependence of the energy of
MO17 is the reason for the change in geometry. MO17 of
ZnH2 in the linear geometry consists of a nonbonding p-orbital
centered on the Zn atom, as depicted on the top left of Figure
8. The bent form contains the same nonbonding p-orbital, but
in this geometry (see top right of Figure 8) the s-orbitals
overlap between the two hydrogen atoms. Arising from this
change is the formation of an H−H bond. Moreover, as there is
also no electron density over the Zn and H atoms, breakage of
the Zn−H bond results. It is this evolution of MO17 which
leads to the photodissociation of ZnH2 and the formation of
atomic zinc and molecular hydrogen.
A similar but more complex behavior was found in
methylzinc hydride as shown in Figure 9. However, just as in
the ZnH2 system, at 90° MO21 of CH3ZnH shows the
reformation of a C−H bond from the hydrogen that was
bonded to the zinc atom. A node is also evident (see top right
of Figure 9) between the zinc atom and the CH3···H moiety.
The angle dependence found in the TD-DFT calculations of
the LUMO energies is identified as responsible for the clean,
concerted dissociation of the VUV bands in the case of both
molecules. This is the classic behavior predicted by a simple
four-valence electron AH2 Walsh-type diagram.
In contrast to the aforementioned behavior, correctly
anticipated by a Walsh diagram, it is noteworthy that the
V. CONCLUSIONS
The absorption spectra of ZnD2 and CH3ZnH have been
recorded for the first time in the vacuum-UV range. These
“inserted” species were produced from the efficient reaction of
excited 1P1 state atomic zinc with the molecular precursors D2
and CH4. Emission of the 1P1 state was completely quenched in
both solid D2 and CH4; however, 3P state luminescence of
atomic zinc was observed in the Zn/CH4 system. The latter
observation points to the involvement of the spin triplet state in
the relaxation of the lower symmetry CH3ZnH system as it
evolves into the C3v ground state geometry. This state appears
to play no role in the stabilization of ZnD2. TD-DFT
calculations provide reliable descriptions of the excited states
responsible for the VUV absorptions of ZnH2 and CH3ZnH;
however, all of the predicted transition energies are underestimates of the observed bands. Large Franck−Condon factors
may partly explain the blue-location of the observed bands
relative to the vertical excitations generated in the TD-DFT
calculations. The observed photodissociation of these molecules has been modeled successfully with Walsh-type diagrams
generated from the DFT calculations.
■
ASSOCIATED CONTENT
S Supporting Information
*
Lineshape analyses of absorption profiles for Zn/D2, Zn/CH4,
and Zn/Ar (Figure S1), a comparison of the VUV absorption
bands produced in Zn/D2 and Zn/CH4 systems (Figure S2), a
comparison of the TD-DFT predictions for the electronic
transitions of CH3ZnH and H−Zn−H (Figure S3), DFT
calculated overlap dependence of the Zn 4p orbital and H atom
1s orbital (Figure S4), comparison of band origins (Table S1),
and band positions extracted by free-fitting Gaussan curves to
the recorded absorption data (Table S2). This material is
available free of charge via the Internet at http://pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail: john.mccaff[email protected].
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS
We would like to acknowledge Dr. Peter Gürtler and Dr. Sven
Petersen for technical assistance with the experimental work
done at the HASYLAB synchrotron. This research was funded
9177
dx.doi.org/10.1021/jp404695a | J. Phys. Chem. A 2013, 117, 9168−9178
The Journal of Physical Chemistry A
Article
in part by the European Union, TMR, “Access to Large Scale
Facilities” Programme. C.H. also gratefully acknowledges
receipt of a Hume Ph.D. studentship, at NUI-Maynooth.
■
(19) Bracken, V. A.; Guertler, P.; McCaffrey, J. G. Spectroscopy and
Photodissociation of Dimethylzinc in Solid Argon. 1. Vacuum UV
Luminescence Detection/Synchrotron Radiation Photolysis. J. Phys.
Chem. A 1997, 101 (51), 9854−9862.
(20) Gimarc, B. M. Molecular Structure and Bonding; Academic Press:
New York, 1979 Chapter 3.
(21) In Figure 9 theoretical predictions for ZnH2 are used in a
comparison with the experimental data recorded for ZnD2. This is
possible because the electronic spectroscopy is not dependent on the
masses of the isotopes.
(22) Presser, N. Spectroscopy of magnesium in hydrogenic matrixes.
Chem. Phys. Lett. 1992, 199 (1−2), 10−16.
(23) Fajardo, M. E. Matrix-isolation spectroscopy of metal atoms
generated by laser ablation. II. The lithium/neon, lithium/deuterium,
and lithium/hydrogen systems. J. Chem. Phys. 1993, 98 (1), 110−18.
(24) Silvera, I. F. The solid molecular hydrogens in the condensed
phase: fundamentals and static properties. Rev. Mod. Phys. 1980, 52 (2,
Pt. 1), 393−452.
(25) Kerins, P. N.; McCaffrey, J. G. A pair potentials study of matrixisolated atomic zinc. I. Excited 1P1 state dynamics in solid Ar. J. Chem.
Phys. 1998, 109 (8), 3131−3136.
(26) Healy, B.; McCaffrey, J. G. Simulation of atomic cadmium
spectroscopy in rare gas solids using pair potentials. J. Phys. Chem. A
2000, 104 (16), 3553−3562.
(27) Umemoto, H.; Tsunashima, S.; Ikeda, H.; Takano, K.;
Kuwahara, K.; Sato, K.; Yokoyama, K.; Misaizu, F.; Fuke, K. Nascent
internal state distributions of ZnH(X2Σ+) produced in the reactions of
Zn(4 1P1) with some alkane hydrocarbons. J. Chem. Phys. 1994, 101
(6), 4803−8.
(28) Casida, M. E.; Jamorski, C.; Casida, K. C.; Salahub, D. R.
Molecular excitation energies to high-lying bound states from timedependent density-functional response theory: characterization and
correction of the time-dependent local density approximation
ionization threshold. J. Chem. Phys. 1998, 108 (11), 4439−4449.
(29) Because the accuracy of the method depends on the exchangecorrelation functional used, TD-DFT may not provide exact transition
energies. As is now well-known, the exchange-correlation functionals
work very well for ground states and for low-lying valence excited
states. However, for high-lying diffuse excited states TD-DFT
employing conventional functionals may be less reliable, and the
excitation energies can be substantially underestimated. This is because
the exchange-correlation potentials generated from these functionals
decay rapidly (exponentially) rather than the gradual (1/r) asymptotic
decay of the true potential.
REFERENCES
(1) Shayesteh, A.; Gordon, I. E.; Appadoo, D. R. T.; Bernath, P. F.
Infrared emission spectra and equilibrium bond lengths of gaseous
ZnH2 and ZnD2. Phys. Chem. Chem. Phys. 2005, 7 (17), 3132−3142.
(2) Sebald, P.; Vennekate, H.; Oswald, R.; Botschwina, P.; Stoll, H. A
theoretical study of ZnH2: a case of very strong Darling-Dennison
resonance. Mol. Phys. 2010, 108 (3−4), 487−499.
(3) Flory, M. A.; Apponi, A. J.; Zack, L. N.; Ziurys, L. M. Activation
of Methane by Zinc: Gas-Phase Synthesis, Structure, and Bonding of
HZnCH3. J. Am. Chem. Soc. 2010, 132 (48), 17186−17192.
(4) Wang, X.; Andrews, L. Infrared Spectra of Zn and Cd Hydride
Molecules and Solids. J. Phys. Chem. A 2004, 108 (50), 11006−11013.
(5) Greene, T. M.; Andrews, L.; Downs, A. J. The Reaction of Zinc,
Cadmium, and Mercury Atoms with Methane: Infrared Spectra of the
Matrix-Isolated Methylmetal Hydrides. J. Am. Chem. Soc. 1995, 117
(31), 8180−7.
(6) Bracken, V. A.; Legay-Sommaire, N.; McCaffrey, J. G.
Spectroscopy and Photodissociation of Dimethylzinc in Solid Argon.
2. FTIR Detection/ArF Laser Photolysis. J. Phys. Chem. A 1997, 101
(51), 9863−9869.
(7) Breckenridge, W. H. Activation of H-H, Si-H, and C-H Bonds by
nsnp Excited States of Metal Atoms. J. Phys. Chem. 1996, 100 (36),
14840−14855.
(8) Breckenridge, W. H.; Wang, J. H. Dynamics of the reactions of
zinc(4s4p3P1) with hydrogen, hydrogen deuteride, and deuterium:
rotational and vibrational quantum-state distributions of zinc hydride
(zinc deuteride) (ZnH (ZnD)) products. J. Chem. Phys. 1987, 87 (5),
2630−7.
(9) Umemoto, H.; Tsunashima, M.; Tsunashima, S.; Ikeda, H.
Nascent vibrational state distributions of ZnH(X2Σ+) and ZnD(X2Σ+)
produced in the reactions of Zn(41P1) with H2 and D2. Laser Chem.
1995, 15 (2−4), 123−9.
(10) Martinez-Magadan, J. M.; Ramirez-Solis, A.; Novaro, O. MCSCF + MR-CI study of the interaction of atomic zinc, and zinc(1+)
and zinc(2+) ions with the hydrogen molecule. Chem. Phys. Lett. 1991,
186 (1), 107−12.
(11) Healy, B.; Kerins, P.; McCaffrey, J. G. Metal atom (Zn, Cd and
Mg) luminescence in solid neon. Low Temp. Phys. 2012, 38 (8), 679−
687.
(12) Bracken, V. A.; Gurtler, P.; McCaffrey, J. G. Luminescence
spectroscopy of atomic zinc in rare-gas solids. I. J. Chem. Phys. 1997,
107 (14), 5290−5299.
(13) Collier, M. A.; McCaffrey, J. G. Luminescence spectroscopy of
matrix-isolated z6P state atomic manganese. J. Chem. Phys. 2005, 122
(18), 184507/1−184507/15.
(14) Frisch, M. J., et al. Gaussian-03, Revision E.01; Gaussian, Inc.:
Wallingford, CT, 2004.
(15) Murray, C.; Dozova, N.; McCaffrey, J. G.; Shafizadeh, N.; Chin,
W.; Broquier, M.; Crepin, C. Visible luminescence spectroscopy of
free-base and zinc phthalocyanines isolated in cryogenic matrixes. Phys.
Chem. Chem. Phys. 2011, 13 (39), 17543−17554.
(16) Alikhani, M. E. Theoretical study of the insertion reaction of
zinc, cadmium, and mercury atoms with methane and silane. Chem.
Phys. Lett. 1999, 313 (3), 608−616.
(17) NIST Atomic Spectra Database, Version 4; National Institute of
Standards and Technology: Gaithersburg, MD, 2013; http://www.nist.
gov/pml/data/asd.cfm (available: May 2013).
(18) Because of the high repetition rate of the SR excitation pulses,
direct measurement of the Zn/CH4 near-UV emission decaytimes was
not possible. Even for “single-bunch” mode operation, when the
repetition rate of the storage ring is reduced to 1.042 MHz, this was
still too high indicating a very long-lived emission. Accordingly it is
inferred that the decay time of the near-UV emissions must be longer
than 10 μs.
9178
dx.doi.org/10.1021/jp404695a | J. Phys. Chem. A 2013, 117, 9168−9178